Meningitis & Ventriculitis
- Meningitis is infection/inflammation of the coverings of the brain; ventriculitis is infection of the lining of the ventricles. Both live on the surfaces, not in the brain meat.
- Uncomplicated meningitis often looks normal on imaging — the diagnosis is made by lumbar puncture, not the scanner. Imaging hunts for complications and rules out reasons not to tap.
- The classic MRI tell is leptomeningeal enhancement that follows the brain's grooves; for ventriculitis it's enhancement and debris lining the ventricles, often with a fluid-debris level in the dependent horns.
- Diffusion restriction is your friend here: pus restricts, so DWI lights up pus in the sulci, ventricles, or a collected pocket of empyema.
Here's the part nobody warns you about: when a patient is rigid-necked, feverish, and clearly miserable with meningitis, you can scan their head and the images can come back looking boringly normal. That's not the scanner failing you. Meningitis is mostly a spinal-fluid diagnosis — the answer is in the tube the team draws during a lumbar puncture, not on the monitor. So why image at all? Because the brain's coverings are crowded real estate, and when an infection sets up shop there, it can cause trouble that the scanner can see. That trouble is the whole reason we're here.
What's actually getting infected
Think of the brain as a delicate cake wrapped in cling film. The thin, gauzy layer hugging every curve and dipping into every groove is the leptomeninges (the pia and arachnoid), and the watery cerebrospinal fluid (CSF) lives in the space between those wraps. Meningitis is infection of those wraps and the fluid between them.
Now picture the hollow chambers deep inside the cake — the ventricles, the brain's internal plumbing where CSF is made and circulates. Ventriculitis (also called ependymitis) is infection of the lining of those chambers. It's the same villain, just gathered in a different room of the house — and it's the more ominous room, because infected CSF sloshing through the ventricles spreads easily and is hard to clear.
Meningitis and encephalitis are often muddled because both make people febrile and confused. The clean distinction: meningitis inflames the coverings, encephalitis inflames the brain itself. They can coexist (meningoencephalitis), but on imaging they point at different places — surfaces versus substance.
Why we scan before we tap
A big practical role of CT here is a gatekeeper one. Before a lumbar puncture, clinicians sometimes want a head CT to make sure there isn't a mass or a lot of swelling that would make removing spinal fluid risky — you don't want to drop the pressure below an expanding lesion. So the first scan is often a normal-looking CT whose entire job was to say "go ahead and tap." That's a win, not a wasted study.
CT can also catch hydrocephalus — ballooning ventricles when infected, sticky CSF gums up the drainage and CSF can't drain or reabsorb the way it should.
What it looks like on MRI
MRI with contrast is where meningitis finally shows its face. The signature finding is leptomeningeal enhancement: after gadolinium, the inflamed, leaky coverings take up contrast and you see bright lines tracing down into the sulci — the grooves of the brain — like someone went over every crease with a highlighter. Normal meninges stay dark, so this groove-following brightness is the tell.
The most sensitive sequence is often FLAIR, where the normally jet-black CSF in the sulci goes abnormally bright because it's now full of protein, cells, and pus instead of clean fluid. Dirty water glows.
Don't confuse leptomeningeal enhancement (thin, follows the sulci into the grooves) with pachymeningeal (dural) enhancement (thick, smooth, hugs the inner skull and crosses over the grooves without dipping in). Pachymeningeal enhancement points more toward things like low CSF pressure, post-surgical change, or dural disease — a different story entirely.
Ventriculitis: the scary version
Ventriculitis earns its own paragraph because it's bad news and has a distinctive look. Hunt for three things: enhancement of the ependymal lining (a thin bright rim outlining the inside of the ventricles), debris layering in the dependent part of the ventricles (usually the back of the lateral ventricles when the patient is lying down — a little fluid-debris level, like sediment settling in a glass), and restricted diffusion of that debris.
That last one is the high-yield trick. On diffusion-weighted imaging, pus restricts — it's thick and cellular, so water can't move freely. Bright pus on DWI inside the ventricles is a near-slam-dunk for ventriculitis. The same physics flags a brain abscess and an empyema (a trapped pocket of pus): both are thick, cellular collections, so both light up on DWI. If diffusion restriction is fuzzy to you, see DWI and ADC physics.
A patient with an external ventricular drain or recent shunt who turns febrile and confused should make you think ventriculitis even before you scan — hardware crossing into the ventricle is a direct route in. On imaging, look hard at the ependymal lining and the dependent horns for that telltale layering debris.
Reading it out
A quick way to keep the surfaces straight:
| Where the infection lives | What to call it | Imaging tell |
|---|---|---|
| Coverings + surface CSF | Meningitis | Sulcal FLAIR brightness; thin leptomeningeal enhancement following grooves |
| Ventricle lining + internal CSF | Ventriculitis | Ependymal enhancement; dependent debris that restricts on DWI |
| Trapped pocket of pus | Empyema/abscess | Rim-enhancing collection that restricts on DWI |
The single thing to walk away with: in meningitis, a normal scan does not clear the patient — the lumbar puncture does. Your job at the workstation is to find the complications (hydrocephalus, infarcts from inflamed vessels, ventriculitis, empyema) that change how worried everyone should be. Find those, and you've earned your keep.